This is the second course from my Computer Vision series. Face Detection and Face Recognition is the most used applications of Computer Vision. Using these techniques, the computer will be able to extract one or more faces in an image or video and then compare it with the existing data to identify the people in that image. Face Detection and Face Recognition is widely used by governments and organizations for surveillance and policing. We are also making use of it daily in many applications like face unlocking of cell phones etc.
Technological interventions in the fields of biometrics and facial recognition have set new innovations in the domain of artificial intelligence. Nowadays, it is implemented in various applications and industry verticals, from unlocking devices to criminal detection. The face recognition technology can be used to identify or authenticate a person. In just a few seconds based on their facial features; therefore gaining better advantage palm print or fingerprint. One of the significant benefits of Facial recognition is that it doesn't need any human interaction.
Artificial Intelligence is something that is ever-evolving in today's tech-heavy generation. Newer technologies are always being developed which incorporate AI to greatly increase quality of life, help in safety and security, boost entertainment, and much more. As artificial intelligence is created with the express intention of being automated, and automating anything that it is applied to, it can be left to discern and learn on its own. One such way that AI is applied to help in security and quality of life, is object recognition. Object recognition is a computer vision technique for identifying objects in images or videos.
Significant advances in deep learning have obtained hallmark accuracy rates for various computer vision applications. However, advances in deep generative models have also led to the generation of very realistic fake content, also known as deepfakes, causing a threat to privacy, democracy, and national security. Most of the current deepfake detection methods are deemed as a binary classification problem in distinguishing authentic images or videos from fake ones using two-class convolutional neural networks (CNNs). These methods are based on detecting visual artifacts, temporal or color inconsistencies produced by deep generative models. However, these methods require a large amount of real and fake data for model training and their performance drops significantly in cross dataset evaluation with samples generated using advanced deepfake generation techniques. In this paper, we thoroughly evaluate the efficacy of deep face recognition in identifying deepfakes, using different loss functions and deepfake generation techniques. Experimental investigations on challenging Celeb-DF and FaceForensics++ deepfake datasets suggest the efficacy of deep face recognition in identifying deepfakes over two-class CNNs and the ocular modality. Reported results suggest a maximum Area Under Curve (AUC) of 0.98 and an Equal Error Rate (EER) of 7.1% in detecting deepfakes using face recognition on the Celeb-DF dataset. This EER is lower by 16.6% compared to the EER obtained for the two-class CNN and the ocular modality on the Celeb-DF dataset. Further on the FaceForensics++ dataset, an AUC of 0.99 and EER of 2.04% were obtained. The use of biometric facial recognition technology has the advantage of bypassing the need for a large amount of fake data for model training and obtaining better generalizability to evolving deepfake creation techniques.
With the help of a facial recognition system, federal agents could capture a man suspected of abuse. The tool detected him in the background of someone else's photo at the gym, in the mirror. So, the agents were able to get to that gym, ask about the man, and eventually capture him. This real-life story, and many others, encourage businesses to benefit from AI services and deploy facial recognition systems. The global facial recognition market size was evaluated at $3.8 billion in 2020 and is expected to reach $8.5 billion in 2025, growing at a CAGR of 17.2%.